JOURNAL OF BACTERIOLOGY, Sept. 1973, p. 949-956 Vol. 115, No. 3Copyright © 1973 American Society for Microbiology Printed in U.S.A.

Cell-Bound Thiaminase I of Bacillusthiaminolyticus

C. COE AGEE', JUDY H. WILKINS2, AND R. L. AIRTH3

Cell Research Institute, University of Texas, Austin, Texas 78712

Received for publication 8 May 1973

The distribution of the extracellular enzyme, thiaminase I, was determined forlogarithmically growing cultures of Bacillus thiaminolyticus. About 60% of theenzyme is associated with the cells throughout the growth cycle. The remainderof the enzyme is in the culture medium. The release of the cell-bound thiaminaseI is examined under a variety of conditions. The rate and extent of release isdependent on the pH and the nature of the incubation solution. The releaseprocess appears to be relatively independent of de novo protein synthesis, energyderived from oxidative phosphorylation, or divalent metal ions. The absence ofcarbon or nitrogen sources has little effect on the release of the enzyme.Cell-bound thiaminase I probably is the immediate precursor for extracellularthiaminase I found in the culture medium. Washed cells continue to releasethiaminase I at the expense of cell-bound enzyme. In addition, purifiedcell-bound thiaminase I is indistinguishable from purified extracellular thiami-nase I by a number of physical and kinetic criteria.

Various laboratories have examined the sitesof synthesis and mechanisms of release of bacte-rial extracellular enzymes (4). Thiaminase I(EC 2.5.1.2), an extracellular enzyme producedin large quantities by Bacillus thiaminolyticus,provides an interesting system for the study ofthis problem. This enzyme catalyzes a base-exchange reaction which results in the destruc-tion of thiamine (Fig. 1). Destruction of thisvital cofactor inside the cell would disruptcellular metabolism. It is therefore likely thatthiaminase I is synthesized either as an inactiveprecursor (eg., zymogen) or within some type ofcellular compartment.

Preliminary experiments with antibody toextracellular thiaminase I indicate the presenceof an enzymatically inactive thiaminase I in thedebris of sonically disrupted B. thiaminolyticuscells (11). This observation therefore supportsthe zymogen theory of thiaminase I origin. Ourattempts to isolate this material lead to thediscovery of a previously unreported activecell-bound thiaminase I.

This communication reports the conditionswhich favor the production of active, cell-bound

' Present address: Department of Biochemistry, Universityof Illinois, Urbana, Illinois 61801.

2Present address: Department of Biochemistry, VirginiaPolytechnic Institute, and State University, Blacksburg,Virginia 24060.

3 Deceased.

thiaminase I and methods which release thisenzyme. Purification and partial characteriza-tion of cell-bound thiaminase I suggests that thecell-bound enzyme is the precursor for theextracellular thiaminase I.

MATERIALS AND METHODSCulturing of the organism. Cultures of B. thiami-

nolyticus, strain M, were grown in a medium consist-ing of 30 mM sucrose, 17 mM sodium citrate, 8 mMKH2PO4, 42 mM Na2HPO4, 50 mM (NH4)2HP04, 0.1mM FeSO4.7H20, 0.05 mM MgSO4.7H20, and 3 x10-8 M thiamine, hydrochloride. The medium (with-out thiamine or magnesium sulfate) was adjusted topH 7.8 prior to autoclaving. Thiamine and magne-sium sulfate were sterilized by filtration (0.22 jm poresize; Millipore Corp.) and added aseptically to thecooled medium. Cultures were grown at 37 C withshaking and then harvested by centrifugation (0-2 C)at 10,000 x g for 20 min.

Release of cell-bound thiaminase I. The cellpellets from two or three, 500-ml, 20-h cultures werewashed once with 20 ml of ice-cold, 0.85% sodiumchloride solution usually containing 100 gg of chlor-amphenicol (CAM) per ml. About 10 to 20% of thetotal cell-bound thiaminase I activity was removed bythis wash. Little additional activity (<5%) was re-moved by repeated washing of the cells under theseconditions. To determine the effects of various envi-ronments on the release of the cell-bound enzyme,cells were incubated in various test solutions consist-ing of buffers, salts, sugars, media, or deionized water.Washed cells were resuspended at approximately 'Voo

949

on July 11, 2018 by guesthttp://jb.asm

.org/D

ownloaded from

AGEE, WILKINS, AND AIRTH

NHa

3AI//E$NH2

N cH BASE'*-Z-7"' 4- CHI-BABE

H1,C4"2H5OH HS " ~ ~ ~ '~iI2H5OH

FIG. 1. Generalized base-exchange reaction cat-alyzed by thiaminase I.

of' their original culture volume in ice-cold, deionizedwater containing 200 gg of CAM per ml. Immediately,a volume of this cell suspension was added to an equalvolume of a test solution. The test solution and cellsuspension were mixed; they were incubated in a

water bath. The f'inal volumes of the incubationmixtures varied from 6 to 30 ml depending on theexperiment. At the end of the incubation period thecells were removed by centrifugation for 1 min at30,000 x g. The supernatant fluids and cell pelletswere then assayed for thiaminase I activity andprotein concentration.

Purification of cell-bound thiaminase I. Prepa-ration ofcrude enzyme. Six liters of culture grown for24 h on defined medium were harvested (0-2 C) bycentrifugation at 10,000 x g. The pelleted cells were

resuspended in 30 ml of 0.05 M sodium citrate bufferat pH 4.7 and incubated for 15 to 20 min at 37 C. Thecells were removed by centrifugation at 30,000 x g for30 min at 0 to 2 C. The supernatant solution contain-ing the thiaminase I activity was decanted into a

Diaflo ultrafiltration cell equipped with a UM-10membrane (Amicon Corp., Lexington, Mass.), dithio-threitol (DTT) was added to a final concentration of10 M, and the supernatant solution was concen-

trated (0-2 C) to 3 or 4 ml under nitrogen gas at 50psi.

Gel filtration. The concentrated supernatant solu-tion was centrifuged (30,000 x g for 30 min) to removeprecipitated proteins. The 30,000 x g supernatantsolution was loaded on a Sephadex G-150 column (2.5by 36 cm) equilibrated in 0.1 M sodium phosphatebuffer at pH 7.0 and eluted at 0 to 2 C with the same

buff'er at a flow rate of' about 10 ml per h. Three-milli-liter fractions were collected and assayed for thiami-nase I activity. Fractions containing thiaminase Iactivity were pooled, DTT was added to a finalconcentration of' 10- 3M, and the enzyme solution wasconcentrated as before.

Preparative gel electrophoresis. A prep-diskpolyacrylamide electrophoresis column (PD-150,Canal Industrial Corp., Rockville, Md.) was preparedcontaining a column (2.5 to 3.0 cm) of a 10% cross-

linked, pH 8.5 tris(hydroxymethyl)aminomethane(Tris)-glycine separating gel. The separating gel con-

tained 10'-4M DTT. The one-centimeter long, 3%stacking gel (pH 6.9) was polymerized with ammoniumpersulfate instead of' riboflavin. The electrode bufferswere 0.03 M Tris-glycine at pH 8.0. The elution bufferwas a pH 9.0, 0.004 M Tris-glycine buffer containing10 ' M DTT. The concentrated enzyme solutionfrom the Sephadex column was adjusted to pH 8 with1 M Tris and then loaded on the prep-disk column.Electrophoresis was performed at a current of' fourmilliamps and 4 C. The proteins were eluted fromthe column at a buffer flow rate of 20 to 30 ml per hand collected as 3-ml fractions. Fractions containingthiaminase I activity were pooled and dialyzed against

several changes of 0.01 M sodium phosphate buffer,pH 7.0, and then stored at 0 to 4 C. Table 1 presentsthe data from a typical purification experiment.

Cell breakage. Cell pellets were resuspended in 0.1M sodium citrate at pH 3.0 or 4.7 and subjected toultrasonic oscillation by using a Branson model S-75sonifer equipped with a microprobe and operating at10,000 cycles/s for a total time of 30 s (six 5-s burstswith 10- to 15-s cooling periods between bursts).Cooling was accomplished with an ice, water, sodiumchloride bath. Cellular debris were removed by cen-trifugation at 30,000 x g for 30 min. This proceduredoes not result in the denaturation of thiaminase I.Enzyme assays. Thiaminase I activity was meas-

ured by either of' two methods. In most cases, thespectrophotometric assay of Douthit and Airth (2) asmodified by Wittliff and Airth (13) was used. Certainspecific instances employ a radiometric assay inwhich the disappearance of thiazole-2- "C-thiamine.hydrochloride (25.2 mCi/mmol) and the appearance ofthiazole-2- "C was monitored. Reactions contained1.15 mM aniline, 0.1 M sodium phosphate buffer, pH5.8, and 39.6 ,M thiazole-2- 4C-thiamine hydrochlo-ride in a total volume of 0.1 ml. Reactions wereprepared and run as described for the spectrophoto-metric assay. Reactions were stopped by the additionof' an equal volume of' saturated trichloroacetic acidsolution and the reactants and products were thenseparated on silica gel G thin-layer chromatographicplates developed with 1-butanol-acetic acid-water(4/1/5). The dried plates were cut at 1-cm intervalsand the powder was suspended in Bray's solution andcounted in a liquid scintillation counter. Heatedenzyme (100 C for 30 min) and controls lackingenzyme were also analyzed. One unit of thiaminase Iactivity was defined as the amount of enzyme re-quired to catalyze the formation of 1 ,umol of productin 1 min at 25 C.

Isocitrate dehydrogenase (EC 1.1.1.41) activity isdetermined by the method of Ochoa (7). Cells for thisassay are sonically disrupted in 1 mM Tris-hydrochlo-ride buffer, pH 8.0. One unit of ICDH activity isdefined as the amount of enzyme which catalyzes thereduction of 1 mol of' nicotinamide adenine dinucleo-tide (NADP) to NADPH2 in 1 min at 37 C.

Protein was estimated by the method of' Lowry et

TABLE 1. Purification of thiaminase I

TotalSpecificVol PoenAtv activityStep (ml (rotmin ity (mnU/ activity(MI(ig/l)ml) (MU) (Mg)

Release super- 32 1,270 970 30,845 765natanta

G-150 pool" 55 112 393 21,615 3,500PD-150 poolc 65 29 177 11,505 6,100

apH 4.7 sodium citrate, 30,000 x g incubationsupernatant fluid.

b Pooled activity fractions from Sephadex G-150column.

c Pooled activity fractions from P-150 prep-diskelectrophoresis column.

950 J. BACTERIOL.

on July 11, 2018 by guesthttp://jb.asm

.org/D

ownloaded from

CELL-BOUND THIAMINASE I OF B. THIAMINOLYTICUS

al. (5) by using bovine serum albumin as a standard.The methods of Shapiro et al. (10) and Ornstein (8)were employed for sodium dodecyl sulfate (SDS)-polyacrylamide electrophoresis and analytical polya-crylamide disk-gel electrophoresis, respectively. Im-munodiffusion was performed by the double-diffusionmethod described by Ouchterlony (9).

Chemicals and organism. Chloramphenicol was agift of H. E. Mechame of Parke, Davis and Co.Thiazole-2- "C-thiamine hydrochloride was pur-chased from Amersham/Searle. B. thiaminolyticusstrain M was originally supplied to us by R. Hayashiof the Department of Microbiology, Yamaguchi Uni-versity School of Medicine, Ube, Japan. Rabbitantibodies against purified extracellular thiaminase Iwere prepared by J. L. Wittliff (14).

RESULTSDouthit and Airth (2) first investigated the

presence of a cell-bound thiaminase I in B.thiaminolyticus. They reported that throughoutthe growth cycle less than 10% of the thiaminaseI activity was cell-bound and greater than 90%of the enzyme occurred in the culture medium.These initial studies involved suspension ofcells in distilled water and complete cellulardisruption by sonic oscillation. Subsequentstudies by Wang et al. (12) and Agee and Airth(1) suggest that these initial studies present afalse picture of the distribution of thiaminase I.Specifically, Agee and Airth (1) have shownthat under certain conditions both cell-boundand extracellular thiaminase I are rapidly inac-tivated in the presence of equimolar quantitiesof thiamine. Wang et al. (12) reported thatcomplete sonic disruption of B. thiaminolyticusreleased about 3 x 10-20 mol of thiamine andthiamine phosphates per cell. Calculationsbased on the data presented here (Fig. 2) showthat cultured under optimal conditions B. thia-minolyticus has about 2.5 x 10-20 mol ofcell-bound thiaminase I per cell. Thus, much ofthe cell-bound enzyme may have been inacti-vated by the intracellular thiamine and thia-mine phosphates released during sonic treat-ment. In addition, the sonic treatment proce-dure used in the earlier studies was too harsh,resulting in some sonic denaturation of thiami-nase I. Full power, continuous sonic treatmentof a cell-free extract of thiaminase I results inthe loss of as much as 30% of the enzyme'sactivity in 60 s. It therefore seemed possiblethat much of the cell-bound enzyme was inacti-vated during the isolation procedure used in theearlier studies. We, therefore, reexamined thedistribution of thiaminase I activity in activelygrowing cultures of B. thiaminolyticus.

If cells are disrupted in pH 3.0 citrate prior toassay, approximately 60% of the total thiami-

61-

IS w,0

wC/) 3z

II.-

0

600

Ji o

300 I0

- 0

18 20 22 24

HOURSFIG. 2. Production of thiaminase I and ICDH dur-

ing the logarithmic phase of growth. Symbols: 0,

cell-bound thiaminase I; 0, extracellular thiaminaseI; 0, intracellular ICDH; U, extracellular ICDH.

nase I activity resides in the cellular fractionthroughout the logarithmic phase of growth(Fig. 2). The observed increase in thiaminase Iactivity per cell is consistent with the datareported earlier for the extracellular enzyme (2).The extracellular enzyme apparently does notarise from cell lysis since isocitrate dehydroge-nase (ICDH), a known intracellular enzyme (7X,is found exclusively inside the cell at all stagesof growth.

Further proof that the thiaminase I activityfound with the cellular fractions is outside thecells' permeability barrier but still firmly at-tached to the cell comes from experiments inwhich cells were incubated with antibody topurified thiamine I. Saline-washed cells were

rapidly agglutinated in 0.85% NaCl in the pres-ence of antithiaminase I antibody. However,after the removal of thiaminase I by incuba-tion for 15 min at 37 C in 0.05 M sodium citratebuffer (pH 4.7) the cells were not agglutinatedby the antibody. Cells incubated for 15 min insodium citrate, pH 4.7 at 0 C, a temperatureat which less than half of the enzyme is releasedfrom the cell, still agglutinated in the presence

I I I I

I I I I

951VOL. 115, 1973

on July 11, 2018 by guesthttp://jb.asm

.org/D

ownloaded from

AGEE, WILKINS, AND AIRTH

of the antibody. Thus cell-bound thiaminase Imust be at or near the cells' surface to be ac-cessible to the antibody and the enzyme mustbe rather tightly bound to the cell or cell ag-glutination would not have occurred.These data suggest two possible relationships

between cell-bound and extracellular thiami-nase I. First, cell-bound and extracellular thia-minase I can be two separate enzymes withsimilar catalytic properties but different func-tional roles. Secondly, the cell-bound enzymemay be the immediate precursor to the extracel-lular enzyme.Two approaches can be used to decide be-

tween these alternatives. First, one can examinethe conditions which affect the release of thecell-bound enzyme. Second, one can purify thecell-bound enzyme to apparent homogenity andcompare its physical and catalytic propertieswith the extracellular enzyme.Table 2 shows that saline-washed cells con-

tinue to release thiaminase I when suspended infresh medium. Cell-bound thiaminase I de-creases as extracellular thiaminase I increasesduring the incubation. Thus, the cell-boundenzyme provides an apparent source for ex-tracellular thiaminase I. Protein synthesis andoxidative phosphorylation inhibitors, or the ab-sence of carbon sources, nitrogen sources, ordivalent cations do not affect the release of theenzyme from the cells.Examination of the release phenomenon

through longer term incubations is impracticalfor under these favorable conditions cell divi-

TABLE 2. Release of cell-bound thiaminase I fromwashed cellsa

Thiaminase I activity(mU/ml)

Incubation medium

bound Released

Complete (minus thiamine-hydrochloride)'

Zero time 53 2830 Min-incubation at 37 C 9 86

- Carbon source 11 105- Nitrogen source 12 89- Carbon and nitrogen 13 116

sources+ 100 gg ofCAM per ml 12 109+ 10-4MDNP 9 103+ 10- 2M EDTA, - Fe2+ 8 110and Mg2+

aCells were washed with salineCAM.

solution without

'Defined medium as described in Materials andMethods.

sion interferes with these assays. Similarly,long-term assays with CAM, dinitrophenol(DNP), or ethylenediaminetetraacetate(EDTA) are also impractical due to cell death.We therefore sought alternate methods forstudying the release of the enzyme. According-ly, we observed the release of the enzyme fromcells suspended at much higher cell densitiesand incubated in various solutions under avariety of conditions.

Figure 3 and Table 3 indicate that the releaseof cell-bound thiaminase I is influenced by thecomposition and pH of the incubation solution.In general, acidic pHs (4.7) or hypotonic condi-tions (deionized water) favor release of theenzyme. Further, the increased release of theenzyme during incubations with acidic buffers(ie., near pH 4) is not reflected in increasedrelease of protein at these acid pHs. Thus, theenzyme released at pH 4 to 6 has a higherspecific activity than the thiaminase I releasedat pH values of 6 and above or 4 and below.

I I I I I I

200 - - 900

150 - 600z

100 300 0w a

Ilx ~~~~~~~~~E

Cl) N...Cr~iio 6OOZ_j

'si

2L_

2~~~~~~~~~~-

2 3 4 5 6 7 8FINAL pH

FIG. 3. Effect of pH on the release of cell-boundthiaminase I. Saline-washed, CAM-treated cells areincubated for 5 min at 37 C in 0.1 M sodiumphosphate buffers. Symbols: 0, thiaminase I activityreleased; *, specific activity of thiaminase I released;or 0.1 M sodium citrate buffers, A, thiaminase Iactivity released; A, specific activity of thiaminase Ireleased. The pH of the medium is measured at theend of the incubation period after the removal of thecells by centrifugation.

952 J . BACTERIOL.

on July 11, 2018 by guesthttp://jb.asm

.org/D

ownloaded from

CELL-BOUND THIAMINASE I OF B. THIAMINOLYTICUS

TABLE 3. Release of cell-bound thiaminase I invarious solutions

Activity Activityreleased removedduring mU by sonic

Incubation solution mU ing mU treat-incuba- ment"otional% (%ofoftotal) total)

0.05 M sodium citrate, 2,475 92 208 8pH 4.7

0.05M sodium citrate, 1,206 41 1,705 59pH 5.9

Deionized waterc 2,454 89 291 110.15M mannitolc 894 35 1,674 650.075 M sodium 645 27 1,716 63

chloridec0.137 M (0.85%) 811 30 1,914 70sodium chloridec

0.15M glucosec 1,071 38 1,706 62

a Saline-washed, CAM-treated cells were sus-pended in the solutions, incubated for 5 min at 37 C,centrifuged, and the supernatant fluids were assayedfor thiaminase I activity.

b At the end of the incubation period, the pelletedcells were sonically treated at pH 4.7 (see Materialsand Methods) and the supernatant fluids were as-sayed for thiaminase I activity.

C"Unbuffered solutions were adjusted to pH 7 withsodium hydroxide prior to the addition of cells.

More importantly, other experiments show thatthe enzyme does not return to the cell uponraising the pH or adding an osmotic support tothe suspension (Tables 4 and 5).The release of cell-bound thiaminase I in

response to the pH of the incubation medium isnot the result of extensive cell lysis. Microscopeexamination of the cells before and after incu-bation does not reveal gross changes in theirmorphology. In addition, viable cell counts onbrain-heart-infusion agar plates indicate that90% of the cells survive 5-min incubations in0.05 M sodium citrate at pH values of 3.8 andabove. Below pH 3.8 the survival rate decreasesrapidly. These e bservations do not reflect apH-mediated activation or inactivation of theenzyme. The activity of a crude, cell-free thia-minase I solution is stable for up to 18 h at roomtemperature at pH values from 2 to 7.

It is probable that the varied extents ofthiaminase I release shown in Table 3 and Fig. 3reflect differences in rates of thiaminase I re-lease rather than the total amount of enzymereleasable. Accordingly, we examined the re-lease of cell-bound thiaminase I as a function ofthe length of the incubation period (Fig. 4). Atzero time, the amount of thiaminase I activityreleased is a function of the pH of the incuba-

tion medium. However, the subsequent rate andextent of release of the enzyme is influenced byboth the pH and the nature of the incubationmedium.These results cannot be due to differential

rates of thiaminase I synthesis during the incu-bations. In all cases the total amount of thiami-nase I activity remained constant throughoutthe incubation period. In addition, the releaseof the enzyme from cells suspended in sodiumcitrate at pH 4.7 is not affected by the additionof inhibitory levels of known protein synthesis,transcription, or oxidative phosphorlation in-hibitors (Table 6). Similar results were obtainedat pH 5.9.

Thus, greater than 90% of cell-bound thiami-nase I is available for release. Further, with theconditions for the release of the cell-boundenzyme firmly documented, a comparison ofextracellular and cell-bound thiaminase I ispossible.The specific procedures that purify extracel-

lular thiaminase I (3, 6, 13) do not yieldsimilarily homogeneous thiaminase I prepara-tions when applied to cell-bound enzyme. Wetherefore devised a purification procedure forcell-bound thiaminase I which yields a single,

TABLE 4. Release of cell-bound thiaminase I

Thiaminase I activity(mU)

Incubation solutiona Cell-bound Extracellu-(released by lar (releasedsonic treat- during in-

ment) cubation)

Sodium citrate-phosphate,pH 4.7b

Zero time 508 72515-Min incubation 15 1,242

Adjusted pH to 7.0 with NaOHZero time after pH adjust- 37 1,198ment

15 Min after pH adjust- 43 1,221ment

Deionized water, pH 7.0Zero time 1,152 8815-Min incubation 246 1,006

Addition NaCl to final concn.of 0.85%

Zero time after NaCl addi- 196 1,029tion

15 Min after NaCl addition 185 1,047

a Saline-washed, CAM-treated cells were sus-pended and incubated at 37 C. At the indicated timessamples were withdrawn, centrifuged, and the super-natant fluids and cells were assayed for thiaminase Iactivity.

5Citric acid (0.05 M) and 0.1 M Na2HPO4 weremixed to a final pH of 4.7.

VOL 115, 1973 953

on July 11, 2018 by guesthttp://jb.asm

.org/D

ownloaded from

AGEE, WILKINS, AND AIRTH

TABLE 5.

Thiaminase I activity (mU)

CellularCells fraction Supernatant

(released by fractionasonic treat- (extracellular)

ment)

Released cells"Zero time 30 030-Min incubation 9 0

Released cells + crudethiaminase I

Zero time 352 2,37415-Min incubation 431 2,37830-Min incubation 442 2,367

aThe supernatant fluid containing the thiaminase Iactivity from the pH 4.7 incubation was dialyzed into0.1 M sodium phosphate, pH 7.0, and then concen-trated fourfold in a Diaflo ultrafiltration apparatus.Precipitated proteins were removed by centrifugation.The "released cells" were resuspended and incubatedin either the concentrated enzyme solution or 0.1 Msodium phosphate buffer, pH 7.0. Samples wereremoved at the indicated times, centrifuged, and thesupernatant fluids and cells were assayed for thiami-nase I activity.

b Saline-washed, CAM-treated cells were freed ofcell-bound thiaminase I activity by a 15-min incuba-tion at 37 C in 0.05 M sodium citrate, pH 4.7, followedby three washes with ice-cold 0.85% sodium chloridesolution.

homogeneous protein on gel electrophoresis inTris-glycine buffer or buffered SDS solution.This same "homogeneous" cell-bound thiami-nase I does reveal slight inhomogenity uponultracentrifugal analysis, but this may alsorepresent polyacrylamide contamination fromthe final step of the purification procedure.

This purification of cell-bound thiaminase I,and the availability of previously purified ex-tracellular thiaminase I allow one to ask ifcell-bound and extracellular thiaminase I arethe same protein. Mixtures of purified cell-bound and extracellular thiaminase I migrate asa single band in pH 8.9 analytical polyacrylam-ide electrophoresis gels. Antibody to purifiedextracellular enzyme gives only one precipitinband in double-diffusion-agar gels when reactedwith cell-bound or extracellular thiaminase Ieither separately or mixed together (Fig. 5). Theobserved ultraviolet light spectrum of the puri-fied cell-bound enzyme and that previously re-ported for extracellular thiaminase I (13) aresimilar with absorption maxima and minima at277 and 252 nm, respectively, and optical den-sity at 280 nm (OD2sonm)/OD2e,onm ratios of

a]__ 0o/LuJ

40:

C/)

z

AHZ 80r04

()40 k-

0 20 40 60MINUTES

FIG. 4. Release of cell-bound thiaminase I as a

function of the incubation time. Saline-washed,CAM-treated cells are suspended in various solutionsand incubated at 37 C. At each time interval, a

sample is removed, centrifuged, and the cells andsupernatant fluid were assayed for thiaminase Iactivity. Symbols: 0, 0.05 M sodium citrate, pH 4.7;0, pH 5.9; A, deionized water; A, 0.15 M mannitol.

TABLE 6. Effects of inhibitors on the release ofthiaminase I from saline washed cellsa

Incubation solution Thiaminase I activityreleased (mU)

0.05 M sodium citrate, pH 4.7, 98zero time

0.05 M sodium citrate, pH 4.7, 18430min

+ 10-4M DNP 168+ 10-3Mpuromycin 190+ 10-' M actinomycin D 174+ 10-3 MCAM 179a Cells were washed, resuspended, and incubated

(at 37 C) as described in the Materials and Methodsexcept that the wash and resuspension solutions didnot contain CAM.

954 J. BACTERIOL.

on July 11, 2018 by guesthttp://jb.asm

.org/D

ownloaded from

CELL-BOUND THIAMINASE I OF B. THIAMINOLYTICUS 955

~~.7 ~ K

,'0eXi,'d':i;':k-)0< 0' )? 000 'ds

....

_Y#./'.4'/4j.**i; ;.'

FIG. 5. Reaction of purified cell-bound and extracellular thiaminase I with rabbit antibodies to purifiedextracellular thiaminase I. (ab, antithiaminase I; 1, 2, 4-purified cell-bound thiaminase I; 3, 5, 6-purifiedextracellular thiaminase I).

VOL. 115, 1973

on July 11, 2018 by guesthttp://jb.asm

.org/D

ownloaded from

AGEE, WILKINS, AND AIRTH

1.65. The observed specific activity of severalof the preparations of purified cell-bound thi-aminase I of 6,000 to 7,000 mU per mg of pro-tein agree well with the 6,100 mU per mg ofprotein value previously reported by Wittliffand Airth (13) for the extracellular enzyme. Bothcell-bound and extracellular thiaminase I dem-onstrate the same 44,000 to 45,000 dalton massupon SDS-polyacrylamide electrophoresis gelsand ultracentrifugation analysis. Lastly, by us-ing the assay conditions of Wittliff and Airth,the Km and Vmax values determined for cell-bound thiaminase I are the same as those pre-viously reported for extracellular thiaminase I(13).

DISCUSSIONTwo lines of evidence suggest that the cell-

bound enzyme is the immediate precursor forextracellular thiaminase I. First, extracellularthiaminase I is produced by washed cells at theexpense of cell-bound enzyme (Tables 2 and 3,and Fig. 4). Secondly, thiaminase I releasedfrom the cell at low pH is indistinguishable byvarious physical and kinetic criteria from ex-tracellular enzyme produced during the culturecycle.The inability of certain inhibitors of protein

synthesis, oxidative phosphorylation, or tran-scription or the absence of carbon sources,nitrogen sources, or divalent cations to affectthe release of cell-bound thiaminase I impliesthat the release of the enzyme is independent ofits synthesis. However, Fig. 2 indicates that theamount of cell-bound thiaminase I per cellreaches a plateau of about 1,000 molecules ofenzyme per cell toward the end of the logarith-mic phase of growth, whereas the amount ofextracellular enzyme continues to rise. Thissuggests that under physiological conditions thesynthesis and release of the enzyme may not betotally independent processes. Specifically, onecan envision that when all of the binding sitesfor thiaminase I are occupied that the rate ofsynthesis influences the rate of release or viseversa. In addition, the effects of pH and osmoticpressure on the release of the enzyme suggestthat cell-bound thiaminase I is located at ornear the surface of the cell, perhaps within amesosome or in the periplasmic space. Both ofthese two locations are subject to alteration bychanges in pH or osmotic pressure.Nothing is known about the physiological role

of thiaminase I in the cell's metabolism, yet acomparison of the effect of pH on the catalytic

activity of the enzyme in the base-exchangereaction to its effect on the release of theenzyme from the cell suggests that it is thecell-bound form of thiaminase I which is physio-logically important to the organism. Thiami-nase I exhibits maximum catalytic activity overa broad range of pH values from around 5.6 and7.0. The catalytic activity falls off rathersharply in both the acidic and basic directions.Little or not activity remains below pH 4.2 orabove pH 8.5 (13). Thus, full catalytic activitycoincides with minimum release from the cell.

ACKNOWLEDGMENTSThis investigation was supported in part by Public Health

Service grant RO1AM-11222 from the National Institute ofArthritis and Metabolic Diseases and by Training Grant inCellular Biology, 5T1 GM-00789, from the National Instituteof General Medical Sciences.

LITERATURE CITED1. Agee, C., and R. L. Airth. 1973. Reversible inactivation of

thiaminase I of B. thiaminolvticus by its substrate,thiamine. J. Bacteriol. 115:957-965.

2. Douthit, H. A., and R. L. Airth. 1966. Thiaminase I ofBacillus thiaminolyticus. Arch. Biochem. Biophys.113:331-337.

3. Ebata, J., and K. Murata. 1961. The purification ofthiaminase I produced by Bacillus thiaminolyticus. J.Vitaminol. 7:115-121.

4. Lampen, J. 0. 1965. Secretion of enzymes by microorga-nisms. Symp. Soc. Gen. Microbiol. 15:115-133.

5. Lowry, 0. H., N. J. Rosenbrough, A. L. Farr, and R. J.Randall. 1951. Protein measurement with the Folinphenol reagent. J. Biol. Chem. 193:265-275.

6. Murata, K. 1965. Thiaminase, p. 220-254. In N. Shima-zono and E. Katsura (ed.), Vitamin B. Research Com-mittee of Japan, review of Japanese literature on beri-beri and thiamine. Igaku Shoin Ltd., Tokyo.

7. Ochoa, S. 1955. Isocitrate dehydrogenase system (TPN)from pig heart, p. 699-704. In S. P. Colowick and S. P.Colowick and N. 0. Kaplan (ed.), Methods of enzy-mology, vol. 1. Academic Press Inc., New York.

8. Ornstein, L. 1964. Disc electrophoresis. I. Backgroundand theory. Ann. N.Y. Acad. Sci. 121:312-349.

9. Ouchterlony, 0. 1949. Antigen-antibody reactions in gel.Acta Pathol. Microbiol. Scand. 26:507-515.

10. Shapiro, A. L., E. Vinuela, and J. V. Maizel, Jr. 1967.Molecular weight estimation of polypeptide chains byelectrophoresis in SDS-polyacrylamide gels. Biochem.Biophys. Res. Commun. 28:815-820.

11. Wang, L., and R. L. Airth. 1967. Repression of thiaminaseI in Bacillus thiaminolyticus. Biochem. Biophys. Res.Commun. 27:325-330.

12. Wang, L., J. H. Wilkins, and R. L. Airth. 1968. Repres-sion of thiaminase I by thiamine and related com-pounds in Bacillus thiaminolyticus. Can. J. Microbiol.14:1143-1147.

13. Wittlift, J. L., and R. L. Airth. 1968. The extracellularthiaminase I ot Bacillus thiaminolyticus. I. Purificationand physiological properties. Biochemistry 7:736-744.

14. Wittliff, J. L., W. J. Mandy, and R. L. Airth. 1968. Theextracellular thiaminase I of Bacillus thiaminolyticus.II. Preparation of antisera and serological properties.Biochemistry 7:2380-2384.

956 J. BACTERIOL

on July 11, 2018 by guesthttp://jb.asm

.org/D

ownloaded from